In questo lavoro si è studiato il comportamento a creep della superlega a base nichel Udimet 720 Li, forgiata isotermicamente, a temperature di interesse applicativo (650-700°C) e con sforzi applicati (450-900MPa) tali da produrre tempi a rottura fino a 5400h. La superlega in esame è utilizzata per la fabbricazione di dischi per turbine a gas. Le curve di creep hanno mostrato un comportamento dipendente dal valore degli sforzi applicati: o A bassi sforzi le curve di creep non presentano alcuno stadio stazionario, ma le curve sono dominate da un lungo stadio accelerante che segue un piccolo, a volte insignificante, stadio primario. Viene mostrato che il lungo stadio accelerante è dovuto all'accumularsi, nel materiale, di un danno correlato alla deformazione da creep. o A sforzi più elevati si ottengono curve apparentemente simili alle precedenti, ma viene mostrato che lo stadio accelerante non è dovuto al crescere di un danno reale nel materiale, più semplicemente è dovuto all'aumentare dello sforzo vero applicato sui campioni con il crescere della deformazione, tipico nelle prove di creep eseguite non a sforzo, ma a carico costante. L'evoluzione della densità delle dislocazioni mobili con la deformazione da creep, permette di razionalizzare i risultati sperimentali.
The creep properties of the isothermally forged Udimet 720 Li, an advanced superalloy for gas turbine disc application, have been investigated under constant and varying load and temperature conditions in the 650-700°C and 450-900MPa temperature/stress field. Figs. 2 display some examples of strain vs time creep curves at the two explored temperatures: the majority of the single creep curves consists of an accelerating stage where the strain rate increases from the minimum creep rate obtained after a short and small primary creep stage. Experimental creep tests on overage specimens (Fig. 5) show the quick creep acceleration after the minimum creep rate is not due to a micro structure degradation related to ageing. The increment of the true applied stress originated by the reduction of the gauge section with strain, as in the constant load creep tests usually happens, is the main cause of experimentally detected accelerated creep in the tests performed at the highest applied stresses. At lower applied stresses the influence of the stress increment with strain is not marked and the accelerated creep strain is mainly attributed to an increment of mobile dislocations with the creep strain and described by the following couple of differential equations: ??=??°(1+W) (1) W=C?? (2) where ? is the instantaneous strain rate, ??° is the extrapolated strain rate at ?=0 corresponding, substantially, to the minimum creep rate, ? is the creep strain and W is the damage producing the tertiary creep acceleration. The damage W is proportional to the creep strain trough a stress/temperature depending constant, C(?,T). The value of the parameters can be easily obtained interpolating the constant load creep curves, in particular the value of the parameter C turned out to be a monotonic decreasing function of applied stress, becoming almost constant and close to zero for tests performed at ?750MPa. The minimum creep rate shows an exponential dependence on the stress. The interpolated values of the parameter C (Fig. 7) and ??° (Fig 3) have been utilised to predict the variable load and creep behaviour by the eqs. 1 and 2. Since C is generally a monotonic decreasing function of the applied stress, the damage (W = C?) is not only a function of the strain, as implicit in the Strain Hardening Model, but it also depends on the stress history. Indeed a creep strain accumulated at high stress with a low value of parameter C, is less damaging compared to the same amount of strain accumulated at lower stress having a higher parameter C. Eq.s 1 and 2 have shown to correctly predict the creep behaviour after a step-like variable load and temperature change (Fig. 10), while the Strain Hardening Model prediction fails completely.
Descrizione del comportamento a creep della superlega Udimet 720 Li
Maldini M;
2005
Abstract
The creep properties of the isothermally forged Udimet 720 Li, an advanced superalloy for gas turbine disc application, have been investigated under constant and varying load and temperature conditions in the 650-700°C and 450-900MPa temperature/stress field. Figs. 2 display some examples of strain vs time creep curves at the two explored temperatures: the majority of the single creep curves consists of an accelerating stage where the strain rate increases from the minimum creep rate obtained after a short and small primary creep stage. Experimental creep tests on overage specimens (Fig. 5) show the quick creep acceleration after the minimum creep rate is not due to a micro structure degradation related to ageing. The increment of the true applied stress originated by the reduction of the gauge section with strain, as in the constant load creep tests usually happens, is the main cause of experimentally detected accelerated creep in the tests performed at the highest applied stresses. At lower applied stresses the influence of the stress increment with strain is not marked and the accelerated creep strain is mainly attributed to an increment of mobile dislocations with the creep strain and described by the following couple of differential equations: ??=??°(1+W) (1) W=C?? (2) where ? is the instantaneous strain rate, ??° is the extrapolated strain rate at ?=0 corresponding, substantially, to the minimum creep rate, ? is the creep strain and W is the damage producing the tertiary creep acceleration. The damage W is proportional to the creep strain trough a stress/temperature depending constant, C(?,T). The value of the parameters can be easily obtained interpolating the constant load creep curves, in particular the value of the parameter C turned out to be a monotonic decreasing function of applied stress, becoming almost constant and close to zero for tests performed at ?750MPa. The minimum creep rate shows an exponential dependence on the stress. The interpolated values of the parameter C (Fig. 7) and ??° (Fig 3) have been utilised to predict the variable load and creep behaviour by the eqs. 1 and 2. Since C is generally a monotonic decreasing function of the applied stress, the damage (W = C?) is not only a function of the strain, as implicit in the Strain Hardening Model, but it also depends on the stress history. Indeed a creep strain accumulated at high stress with a low value of parameter C, is less damaging compared to the same amount of strain accumulated at lower stress having a higher parameter C. Eq.s 1 and 2 have shown to correctly predict the creep behaviour after a step-like variable load and temperature change (Fig. 10), while the Strain Hardening Model prediction fails completely.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
